Re: [PATCH -next] arch_topology: Fix cache attributes detection in the CPU hotplug path

From: Conor.Dooley
Date: Thu Jul 14 2022 - 12:10:45 EST

On 14/07/2022 17:00, Sudeep Holla wrote:
> EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe
> On Thu, Jul 14, 2022 at 03:27:09PM +0000, Conor.Dooley@xxxxxxxxxxxxx wrote:
>> On 14/07/2022 16:01, Sudeep Holla wrote:
>>> Interesting, need to check if it is not in atomic context on arm64.
>>> Wonder if some configs are disabled and making this bug hidden. Let me
>>> check.
> OK, it turns I didn't have necessary config options enabled. Enabling
> them, I did see the BUG splat and changing allocation to GFP_ATOMIC
> fixed the same. Can you try that please so that you can test if other
> things are fine.
>>> One possible solution is to add GFP_ATOMIC to the allocation but I want
>>> to make sure if it is legal to be in atomic context when calling
>>> update_siblings_masks.
> So I take is as legal and needs to be fixed to push my patch.

With the GFP_ATOMIC, behaviour is the same as before for me.

Therefore, with the following diff & for RISC-V/DT:

Reviewed-by: Conor Dooley <conor.dooley@xxxxxxxxxxxxx>

diff --git a/drivers/base/cacheinfo.c b/drivers/base/cacheinfo.c
index 65d566ff24c4..4b5cd08c5a65 100644
--- a/drivers/base/cacheinfo.c
+++ b/drivers/base/cacheinfo.c
@@ -356,7 +356,7 @@ int detect_cache_attributes(unsigned int cpu)
return -ENOENT;

per_cpu_cacheinfo(cpu) = kcalloc(cache_leaves(cpu),
- sizeof(struct cacheinfo), GFP_KERNEL);
+ sizeof(struct cacheinfo), GFP_ATOMIC);
if (per_cpu_cacheinfo(cpu) == NULL) {
cache_leaves(cpu) = 0;
return -ENOMEM;